Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Rook + Pawn vs. Rook + Pain in 6-piece

Author: Marc Bourzutschky

Date: 08:46:03 05/16/04

Go up one level in this thread


On May 16, 2004 at 09:10:27, Ed Trice wrote:

>Does anyone know how far away the R+P vs. R+P subset of the 6-piece database is
>from completion?
>
>Can this subset be probed "stand alone" with its subsets (i.e. R + P vs. R, K +
>R vs. K, but not supersets like R + Q vs. R + Q) without any negative impact?

RPxRP has been constructed (I, for one, generated it about 1.5 years ago using
the Chessmaster FEG format) but I'm not aware of any downloadable versions.

Most tablebases contain strictly the material specified, so none of the subgames
obtained via capture or promotion are in the same tablebase (one exception are
the ones generated by Ken Thompson, which contain subgames arrived at through
captures).

Any tablebase position can always be probed "stand alone". However, if you also
want to resolve all the moves available in the position that lead to a different
material configuration, it is possible that you may not have enough information
to find the best move unless you have the tablebase for this new configuration.
This usually not a big issue, but some chess engines don't deal gracefully with
missing subgames.

-Marc



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.